Uncertainty quantification
Document Actions
The goal of uncertainty quantification is to assign an appropriate mathematical model to real-world information with respect to objective and subjective uncertainty.
The choice of an appropriate uncertainty model primarily depends on the characteristics of the available information. That is, the underlying reality with the sources of the uncertainty dictates the model. In engineering, information can be, for example, objective, subjective, dubious, incomplete, fragmentary, imprecise, fluctuating, linguistic, data-based, or expert-specified. In each particular case this information must be analyzed and classified to be eligible for quantification. In general, one of the following three major uncertainty models provides a suitable basis.
Data-based information which is characterized by random fluctuations may be described with the aid of a traditional probabilistic model.
The uncertainty model fuzziness lends itself to describing imprecise, subjective, linguistic, and expert-specified information.
The uncertainty model fuzzy randomness is particularly suitable for adequately quantifying uncertainty that comprises only some (incomplete, fragmentary) objective, data-based, randomly fluctuating information, which can simultaneously be dubious or imprecise and may additionally be amended by subjective, linguistic, expert-specified evaluations.
© Institute of Statics and Dynamics of Structures (TU Dresden)
